Allocation of English Remote Guiding based on Deep Reinforcement Learning and Multi-Objective Optimization

Distance education and traditional education are two teaching methods coexisting in modern society, which play an important role in the process of education popularization. The most obvious advantage of distance learning is that students can not be constrained by time, place, space and other aspects. Through distance learning, students can freely arrange their own learning time and schedule. According to the characteristics of superparameter optimization process, this paper uses deep neural network to build an agent to optimize superparameter. On the basis of priority experience playback, the priority of the prestate of the updated sample is adjusted, which makes the sampling more efficient.